## REPLICATION ARCHIVE: 

Katsumata, Hiroto and Shunya Noda. 2023. "'Kick Them Out' as a Voting Strategy: Theory and Evidence from Multi-member District Elections." Journal of Politics, forthcoming.



# CONTENTS

- This replication archive contains the following folders and files needed to replicate the results in the article:


(1) README.txt: this document.


(2) Dataset

- All datasets are included in "data" folder except for ones with restrictions of redistribution.
- See "HOW TO REPRODUCE THE RESULTS" section for details.

- Included
1. "data/Data_DistrictMagnitudeAugust2016.csv": Results of the experiment conducted by Hix et al. (2017)
2. "data/rou2.csv": Survey data for the 2004 Romanian CLPR election
3. "data/rou2004seat.csv": Seat allocation in the 2004 Romanian CLPR election
4. "data/rou2004share.csv": Additional vote share data in the 2004 Romanian CLPR election
5. "data/Originalcodejes3_1.csv": Code for identifying each candidate for the 2001 Japanese SNTV election
6. "data/Originalcodejes4_5.csv": Code for identifying each candidate for the 2010 Japanese SNTV election

- Not included because of restrictions of redistribution (These datasets should be saved in "data" folder)
7. "cses2.csv": Original dataset of the survey data for the 2004 Romanian CLPR election (Available at: https://cses.org/data-download/cses-module-2-2001-2006/)
8. "jes3.csv": Survey data for the 2001 Japanese SNTV election (Available at: https://ssjda.iss.u-tokyo.ac.jp/Direct/gaiyo.php?lang=eng&eid=0530)
9. "jes4_5.csv": Survey data for the 2010 Japanese SNTV election (Available at: https://ssjda.iss.u-tokyo.ac.jp/Direct/gaiyo.php?lang=eng&eid=0999)
10. "jes4_6.csv": Survey data for the 2010 Japanese SNTV election (Available at: https://ssjda.iss.u-tokyo.ac.jp/Direct/gaiyo.php?lang=eng&eid=0999)


(3) Code

- All codes are included in "scripts" folder except for "functions/functions.R", which is included in "functions" folder

1. "scripts/kickout_experiment.R": for the UK experiment data reanalysis (Tables 1, C.1, D.1--3, and E.1--4; and Figures 3, 4, and C.1)
2. "scripts/kickout_romania.R": for the Romanian CLPR election analysis (Tables C.2 and D.4--9; and Figures 5, 6, and C.2)
3. "scripts/kickout_japan.R": for the Japanese SNTV elections analysis (Tables A.1; and Figures A.1--3)
4. "functions/kickout_functions.R": R functions for the codes 1--3 above
5. "scripts/kickout_romania_data.R": for preparing the dataset ("data/rou2.csv") for "scripts/kickout_romania.R"


(4) Codebook
- "kickout_codebook.pdf"



# HOW TO REPRODUCE THE RESULTS

- Open an R project "kickout.Rproj".
- Run the following codes in the R project.
	- "scripts/kickout_experiment.R": For Tables 1, C.1, D.1--3, and E.1--4; and Figures 3, 4, and C.1.
	- "scripts/kickout_romania.R": For Tables C.2 and D.4--9; and Figures 5, 6, and C.2.
- Download the following datasets and save them in "data" folder.
	- "jes3.csv": (Available at: https://ssjda.iss.u-tokyo.ac.jp/Direct/gaiyo.php?lang=eng&eid=0530)
	- "jes4_5.csv": (Available at: https://ssjda.iss.u-tokyo.ac.jp/Direct/gaiyo.php?lang=eng&eid=0999)
	- "jes4_6.csv": (Available at: https://ssjda.iss.u-tokyo.ac.jp/Direct/gaiyo.php?lang=eng&eid=0999)
- Run the following codes in the R project.
	- "scripts/kickout_japan.R": For Tables A.1; and Figures A.1--3.
- All figures are saved in "figures" folder
- All tables are saved in "tables" folder in a LaTeX format.

- If you want to use the original CSES dataset for the replication, you can do it in the following manner
- Downlad the following dataset in the CSV format and save it in "data" folder.
	- "cses2.csv": (Available at: https://cses.org/data-download/cses-module-2-2001-2006/)
- Run the following codes in the R project.
	- "scripts/kickout_romania_data.R": To make the dataset for the Romanian CLPR election analysis and save it as "data/rou2.csv" in "data"
 folder.
	- "scripts/kickout_romania.R": For Tables C.2 and D.4--9; and Figures 5, 6, and C.2.



# NOTES ON THE FIGURES

- Figure 1 and 2 are just calculated mathematically and not produced either through statistical nor simulation analysis. Thus, they are not reproduced by the codes above.



# SOFTWARE DEPENDENCIES

- All analyses were produced with the following software, operating system environments, and installed packages:

- Operating system: macOS Ventura 13.2.1
- R version 4.0.2
- R packages
	- Hmisc (version 4.4-1)
	- MASS (version 7.3-51.6)
	- sandwich (version 2.5-1)
	- tidyverse (version 1.3.0)
